Suicide research to date has been handicapped by the lack of a uniform, coherent nomenclature and system of classification for describing suicidal behaviors. Our research plan is designed to refine and expand the Preliminary Schema for Classification of Suicidal Behaviors developed by the Center for Studies of Suicide Prevention of the National Institute of Mental Health. This Schema divides suicidal behaviors into three categories: Completed Suicides, Attempted Suicides, and Suicidal Ideation. Five hundred patients admitted to Philadelphia General Hospital and the University of Pennsylvania Hospital because of suicide attempts or threats have been evaluated by members of our research team and classified according to the CSSP Schema. After discharge, a social worker remains in contact with the patient and will record any instances of suicidal behavior occuring during the follow-up period. A Suicidal Intent Schedule and a Medical Lethality Scale are used in the research. Follow-up data will enable us to judge the value of these instruments as predictors of future suicidal behavior. The Depression Inventory and the Hopelessness Scale will also enable us to study the relationship of suicide to depression and hopelessness, and the value of measures of depression and hopelessness as predictors of future suicidal behavior. Further, we have studied learned helplessness in animals and humans. We are currently applying the principle derived from our studies to an outcome study of depressed outpatients. Patients are randomly assigned to either a short term structured psychotherapy group or to a pharmacotherapy group.